The deployment of autonomous systems in the real world, such as self-driving vehicles and unmanned aerial systems, requires rigorous methods for verifying and reasoning about whether a system can maintain safety, broadly understood as the ability to enforce constraints on system behaviors. Safety is one of the most fundamental challenges in deploying autonomous systems in the real world. A diverse landscape of methods has emerged to address this challenge, drawing on techniques from control theory, optimization, and formal verification.

Among these, Control Barrier Functions (CBFs) and Hamilton-Jacobi (HJ) Reachability Analysis have become two prominent families: while they share the overarching goal of performing safety analysis for control systems, they are rooted in distinct mathematical foundations, and hence offer different insights and lend themselves to different analysis and design tools. They also share common open challenges, such as scalability to high-dimensional systems and integration with learned components.

Beyond these two frameworks, perspectives from formal methods offer complementary tools for specifying and verifying safety properties, and increasingly intersect with control-theoretic approaches. Finally, recent advances in machine learning-based methods, such as physics-informed machine learning and reinforcement learning, emerge as effective tools for constructing the safety certificates or conducting the safety analysis.

This workshop aims to bring together researchers working across this broader space—centered on CBFs and HJ reachability, while engaging with adjacent perspectives from formal methods, and bringing in machine learning techniques for advancing design and analysis methods—to exchange ideas and accelerate progress towards the safe deployment of autonomous systems.